Fault Tree Analysis

Fault tree analysis [FTA] is one of the core analyses. Other analyses may require or depend on the fault-tree analysis results or constructs. Fault trees employ various types of gates (Boolean connectives) and events to represent Boolean formulas and to model systems for analysis. SCRAM automatically discovers top-events (graph roots) and runs all requested analysis kinds with those events. The following is a description of some analysis goals and their respective control parameters (knobs).

  1. Find minimal cut sets or prime implicants. Probability input is optional
    • Cut-off probability for products. Unused
    • Maximum order for products for faster calculations.
  2. Find the total probability of a top event and importance values for basic events. Only if probability input is provided
    • Cut-off probability for products. Unused
    • The rare event or MCUB approximation. Optional
    • Mission time that is used to calculate probabilities.

Supported Gate and Event Types

All the event types and Boolean connectives described in the Open-PSA [MEF] are supported. In addition, SCRAM currently “abuses” the Open-PSA [MEF] attributes mechanism to add inessential (for analysis) information or flavors to events. This approach still ensures that the input file is portable across tools.


Add this XML line to a basic event description: <attributes> <attribute name="flavor" value="undeveloped"/> </attributes>